In helping dozens of organizations build on-premises AI initiatives, we have seen three fundamental stages organizations go through on their journey to enterprise-scale AI. First, individual data scientists experiment on proof of concept projects which may be promising. These PoCs then often hit knowledge, data management and infrastructure performance obstacles that keep them from proceeding to the second stage to deliver optimized and trained models quickly enough to deliver value to the organization. Moving to the third and final stage of AI adoption, where AI is integrated across multiple lines of business and requires enterprise-scale infrastructure, presents significant integration, security and support challenges. Today IBM introduced IBM PowerAI Enterprise and an on-premises AI infrastructure reference architecture to help organizations jump-start AI and deep learning projects, and to remove the obstacles to moving from experimentation to production and ultimately to enterprise-scale AI.

As a fan of poker, I know there's no such thing as a sure bet. But randomly pick 100 tech startups and I'd confidently wager that the vast majority are t weaving AI heavily into their narrative. AI hype has become intense, and understandably so. The technology has become part of everyday life, whether it's Netflix predicting what shows we might like based on previous choices, Google's search results consistently improving based on millions of people's clicks, or conversational AI systems like Amazon's Alexa getting to know you. Meanwhile, virtually invisible to mainstream consumers, the business world is agog over harnessing AI to solve a range of problems.

One of the more complex and misunderstood topics making headlines lately is artificial intelligence. People like Elon Musk warn that robots could one day destroy us all, while other experts claim that we're on the brink of an AI winter and the technology is going nowhere. Making heads or tails of it all is difficult, but the best place to start is with deep learning. Here's what you need to know. Artificial intelligence has become a focal point for the global tech community thanks to the rise of deep learning.

Recent advances in graph convolutional networks have significantly improved the performance of chemical predictions, raising a new research question: "how do we explain the predictions of graph convolutional networks?" A possible approach to answer this question is to visualize evidence substructures responsible for the predictions. For chemical property prediction tasks, the sample size of the training data is often small and/or a label imbalance problem occurs, where a few samples belong to a single class and the majority of samples belong to the other classes. This can lead to uncertainty related to the learned parameters of the machine learning model. To address this uncertainty, we propose BayesGrad, utilizing the Bayesian predictive distribution, to define the importance of each node in an input graph, which is computed efficiently using the dropout technique. We demonstrate that BayesGrad successfully visualizes the substructures responsible for the label prediction in the artificial experiment, even when the sample size is small. Furthermore, we use a real dataset to evaluate the effectiveness of the visualization. The basic idea of BayesGrad is not limited to graph-structured data and can be applied to other data types.

Over past several years, deep learning has achieved huge successes in various applications. However, such a data-driven approach is often criticized for lack of interpretability. Recently, we proposed artificial quadratic neural networks consisting of second-order neurons in potentially many layers. In each second-order neuron, a quadratic function is used in the place of the inner product in a traditional neuron, and then undergoes a nonlinear activation. With a single second-order neuron, any fuzzy logic operation, such as XOR, can be implemented. In this sense, any deep network constructed with quadratic neurons can be interpreted as a deep fuzzy logic system. Since traditional neural networks and second-order counterparts can represent each other and fuzzy logic operations are naturally implemented in second-order neural networks, it is plausible to explain how a deep neural network works with a second-order network as the system model. In this paper, we generalize and categorize fuzzy logic operations implementable with individual second-order neurons, and then perform statistical/information theoretic analyses of exemplary quadratic neural networks.

In this paper, we study the problem of data augmentation for language understanding in task-oriented dialogue system. In contrast to previous work which augments an utterance without considering its relation with other utterances, we propose a sequence-to-sequence generation based data augmentation framework that leverages one utterance's same semantic alternatives in the training data. A novel diversity rank is incorporated into the utterance representation to make the model produce diverse utterances and these diversely augmented utterances help to improve the language understanding module. Experimental results on the Airline Travel Information System dataset and a newly created semantic frame annotation on Stanford Multi-turn, Multidomain Dialogue Dataset show that our framework achieves significant improvements of 6.38 and 10.04 F-scores respectively when only a training set of hundreds utterances is represented. Case studies also confirm that our method generates diverse utterances.

We present a transition-based parser that jointly produces syntactic and semantic dependencies. It learns a representation of the entire algorithm state, using stack long short-term memories. Our greedy inference algorithm has linear time, including feature extraction. On the CoNLL 2008--9 English shared tasks, we obtain the best published parsing performance among models that jointly learn syntax and semantics.

Deep neural networks are central to many recent successes of machine learning, including applications such as computer vision, natural language processing, or reinforcement learning. A common trend in deep learning has been to construct larger and deeper networks, starting from the pioneer convolutional network LeNet [19], to networks with tens of layers such as AlexNet [17] or VGG-Net [28], or recent architectures like GoogLeNet/Inception [30] or ResNet [13, 14], which may contain hundreds or thousands of layers. A typical observation is that deeper networks offer better performance. This phenomenon, at least on the training set, supports the intuition that a deeper network should have more capacity to approximate the target function, and leads to a question that has received increasing interest in the theory of deep learning: can all functions that we may care about be approximated well by a sufficiently large and deep network? In this work, we address this important question for the popular ResNet architecture.

A neural network (NN) is a parameterised function that can be tuned via gradient descent to approximate a labelled collection of data with high precision. A Gaussian process (GP), on the other hand, is a probabilistic model that defines a distribution over possible functions, and is updated in light of data via the rules of probabilistic inference. GPs are probabilistic, data-efficient and flexible, however they are also computationally intensive and thus limited in their applicability. We introduce a class of neural latent variable models which we call Neural Processes (NPs), combining the best of both worlds. Like GPs, NPs define distributions over functions, are capable of rapid adaptation to new observations, and can estimate the uncertainty in their predictions. Like NNs, NPs are computationally efficient during training and evaluation but also learn to adapt their priors to data. We demonstrate the performance of NPs on a range of learning tasks, including regression and optimisation, and compare and contrast with related models in the literature.

Deep neural networks excel at function approximation, yet they are typically trained from scratch for each new function. On the other hand, Bayesian methods, such as Gaussian Processes (GPs), exploit prior knowledge to quickly infer the shape of a new function at test time. Yet GPs are computationally expensive, and it can be hard to design appropriate priors. In this paper we propose a family of neural models, Conditional Neural Processes (CNPs), that combine the benefits of both. CNPs are inspired by the flexibility of stochastic processes such as GPs, but are structured as neural networks and trained via gradient descent. CNPs make accurate predictions after observing only a handful of training data points, yet scale to complex functions and large datasets. We demonstrate the performance and versatility of the approach on a range of canonical machine learning tasks, including regression, classification and image completion.